Abstract
Purpose:
Retinal toxicity is a significant cause of progression failure in the drug development process. The aim of these experiments was to develop and characterise a human retinal culture model for assessment of retinal toxicity.
Methods:
The retina was dissected from donor human eyes. Five 4mm diameter explants from the paramacular region and one macula explant was obtained from each retina. Planar sectioning was carried out in which the retinal explant was flat-mounted and cryosectioned (20um slices) in the plane of the nuclear layers. QRT-PCR was carried out (n=4) to assess expression of the retinal cell marker genes RCVRN, RBFOX3, THY1, CHAT, CALB1, PRKCA, RLBP1, GLUL, GFAP. Statistical analysis was carried out using ANOVA.
Results:
Explants taken from the paramacular retina showed no significant difference in expression of any of the retinal cell marker genes indicating consistency across these 5 regions (p>0.05). In comparison, there was a significant decrease in RBFOX3 of 25.6 ± 4.3% and THY1 of 28.7 ± 2.7% and a significant increase in PRKCA of 241.5 ± 14.9% when comparing paramacular to macular samples, reflecting differences in density of retinal ganglion cells and rod ON-bipolar cells between these regions. Gene profiling of retinal cell markers displayed peak expression consistent with the positioning of cells within the retina, with peak expression of RCVRN within the outermost layer, markers for bipolar (PRKCA), horizontal (CALB1) and amacrine (CHAT) cells as well as Müller cell markers (RLBP1 and GLUL) in the regions corresponding to the inner nuclear layer. RBFOX3, THY1 and GFAP showed peak expression in the inner layers of the retina corresponding to the ganglion cell layer and the nerve fibre layer.
Conclusions:
Expression equivalence between paramacular samples enables them to be used in ex vivo explant culture experiments to compare the effects of potential retinotoxic compounds on the human retina. Planar sectioning enables profiling of genes of interest relating to retinotoxic compounds and could also be useful to investigate loss of retinal cell markers following retinotoxic insult in an ex vivo culture model. The human retinal explant model could be of benefit in linking the species gap between in vivo animal and human clinical studies.